Responsibilities
- Design and sustain security measures throughout the entire machine learning lifecycle, including data input, feature processing, model training, storage, deployment, and runtime operations.
- Safeguard Vertex AI components such as pipelines, notebooks, training tasks, model outputs, and endpoints as they transition from experimental to production environments.
- Enforce strict segregation between development, training, staging, and production environments using proper identity management, authorization protocols, and access restrictions.
- Create protective frameworks for autonomous and automated AI systems, defining operational limits, tool access rules, and management of non-human identities.
- Ensure the integrity of the machine learning supply chain by securing feature pipelines, verifying training data lineage, protecting model artifacts, and safeguarding model serving endpoints.
- Incorporate machine learning security practices into existing application security CI/CD workflows, secure software development lifecycle procedures, and security testing infrastructures.
- Adapt application security standards—including identity and access management, secrets handling, and API protection—to machine learning workloads.
- Develop standardized, reusable security designs, reference architectures, and governance frameworks for AI and ML initiatives.
- Deploy monitoring systems to detect AI-specific threats and respond to security incidents involving AI pipelines, model behavior, or agent activities.
- Support the creation of enterprise-wide AI security policies and sustainable operational models for long-term AI governance.
Work Arrangement
Hybrid
